The impact of nonsteroidal anti-inflammatory drugs on the equine gastric microbiome

Abstract

Digesta retention time within specific segments of the equine gastrointestinal tract (GIT) may be more relevant to scientific inquiries than total tract mean retention time (TTMRT); however, measuring retention time in individual segments requires access to the digestive tract. The objective of this study was to compare prececal, cecal, and colonic mean retention time (MRT) with model-derived compartment MRT. A cecally fistulated gelding was fed indigestible particulate and liquid markers to determine TTMRT and dosed with different pairs of particulate and liquid markers into the cecum (hindgut MRT) and into the right ventral colon through the cecocolic orifice (colon MRT). Fecal marker concentrations were fit to stochastic and mechanistic models using nonlinear least squares methods (MATLAB). Total tract MRT and MRT for each GIT segment were compared with model-derived compartment MRT using paired t-test to determine differences and two one-sided tests to determine equivalence. All models resulted in parameter estimates and an acceptable fit to fecal marker excretion curves, but some parameter estimates did not differ from zero (95% CI included 0). Model-derived TTMRT were equivalent (P < .05) to arithmetically calculated MRT. Most GIT segment MRT differed (P < .05) from model-derived compartment MRT. Differences ranged from −26.1 hours to 25.8 hours. In these exploratory data, model derived compartment retention times failed to pair with MRT in different GIT segments. Significant methodological and analytical challenges remain to describe retention time in individual segments of the equine GIT.